[Bug] Passing `trust_remote_code=True` will be mandatory to load this dataset from the next major release of `datasets`.

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• The bug has not been fixed in the latest version.

Type

I have modified the code (config is not considered code), or I'm working on my own tasks/models/datasets.

Environment

{'CUDA available': True, “CUDA available”：True，
'CUDA_HOME': None, 'CUDA_HOME'：无，
'GCC': 'gcc (Ubuntu 11.4.0-1ubuntu122.04) 11.4.0',
'GCC'：'GCC(Ubuntu 11.4.0-1ubuntu122.04)11.4.0',
'GPU 0': 'NVIDIA GeForce RTX 4090 D',
'GPU 0'：'NVIDIA GeForce RTX 4090 D'，
'MMEngine': '0.10.4', 'MMCengine'：'0.10.4'，
'MUSA available': False,
'MUSA available'：False，
'OpenCV': '4.9.0', 'OpenCV'：'4.9.0'，
'PyTorch': '2.3.0', 'PyTorch'：' 2.3.0'，
'PyTorch compiling details': 'PyTorch built with:\n'
'PyTorch编译详细信息'：'PyTorch构建时使用：\n'
' - GCC 9.3\n'
GCC 9.3\n的
' - C++ Version: 201703\n'
C++版本：201703\n
' - Intel(R) oneAPI Math Kernel Library Version '
' -英特尔（R）oneAPI数学内核库版本'
'2023.1-Product Build 20230303 for Intel(R) 64 '
'2023.1-英特尔® 64产品构建20230303'
'architecture applications\n'
'体系结构应用程序\n'
' - Intel(R) MKL-DNN v3.3.6 (Git Hash '
' -英特尔（R）MKL-DNN v3.3.6（Git Hash '
'86e6af5974177e513fd3fee58425e1063e7f1361)\n'
'86e6af5974177e513fd3fee58425e1063e7f1361）\n'
' - OpenMP 201511 (a.k.a. OpenMP 4.5)\n'
'—OpenMP 201511（a.k.a. OpenMP 4.5）\n '
' - LAPACK is enabled (usually provided by '
' - LAPACK已启用（通常由'
'MKL)\n' 'MKL）\n'
' - NNPACK is enabled\n'
'—NNPACK已启用\n'
' - CPU capability usage: AVX2\n'
'—CPU能力使用率：AVX2\n'
' - CUDA Runtime 12.1\n'
' - CUDA 12.1\n'
' - NVCC architecture flags: '
' - NVCC体系结构标志：'
'-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90\n'
'-gencode; arch= compute_50，code =sm_50;-gencode;arch= compute_60，code =sm_60;-gencode;arch= compute_61，code =sm_61;-gencode;arch= compute_70，code =sm_70;-gencode;arch= compute_75，code =sm_75;-gencode;arch= compute_80，code =sm_80;-gencode;arch= compute_86，code =sm_86;-gencode;arch= compute_90，code =sm_90\n'
' - CuDNN 8.9.2\n'
' - Magma 2.6.1\n'
' - Build settings: BLAS_INFO=mkl, '
' -构建设置：BLAS_INFO=mkl，'
'BUILD_TYPE=Release, CUDA_VERSION=12.1, '
'BUILD_TYPE=Release，CUDA_VERSION=12.1，'
'CUDNN_VERSION=8.9.2, ' 'CUDNN_VERSION=8.9.2，'
'CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, '
'CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++，'
'CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 '
'-fabi-version=11 -fvisibility-inlines-hidden '
'-fabi-version=11 -fabability-inlines-hidden '
'-DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO '
'-DLIBKINETO_NOROCTRACER -DUSE_FBGEMM '
'-DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK '
'-DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE '
'-DUSE_XNNPACK -DSYMBOLICATE_移动的_DEBUG_HANDLE '
'-O2 -fPIC -Wall -Wextra -Werror=return-type '
'-O2 -fPIC -Wall -Wextra -Werror=返回类型'
'-Werror=non-virtual-dtor -Werror=bool-operation '
'-Werror=non-virtual-dtor -Werror= bool-operation'
'-Wnarrowing -Wno-missing-field-initializers '
'-Wno-type-limits -Wno-array-bounds '
'-Wno-unknown-pragmas -Wno-unused-parameter '
'-Wno-unused-function -Wno-unused-result '
'-Wno-strict-overflow -Wno-strict-aliasing '
'-Wno-stringop-overflow -Wsuggest-override '
'-Wno-stringop-overflow -Wrest-override'
'-Wno-psabi -Wno-error=pedantic '
'-Wno-psabi -Wno-error=迂腐'
'-Wno-error=old-style-cast -Wno-missing-braces '
'-fdiagnostics-color=always -faligned-new '
'-Wno-unused-but-set-variable '
'-Wno-maybe-uninitialized -fno-math-errno '
'-fno-trapping-math -Werror=format '
'-Wno-stringop-overflow, LAPACK_INFO=mkl, '
'-Wno-stringop-overflow，LAPACK_INFO=mkl，'
'PERF_WITH_AVX=1, PERF_WITH_AVX2=1, '
PERF_WITH_AVX = 1，PERF_WITH_AVX2 = 1，
'PERF_WITH_AVX512=1, TORCH_VERSION=2.3.0, '
PERF_WITH_AVX512 = 1，TORCH_VERSION = 2.3.0，这是怎么回事？
'USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, '
USE_CUDA = ON，USE_CUDNN = ON，USE_CUPARSELT = 1，USE_CUDNN = ON，USE_CUDA = ON，USE_CUDNN = ON
'USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, '
'USE_EXCEPTION_PTR=1，USE_GFLAGS=OFF，'
'USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, '
USE_GLOG = OFF，USE_GLOO = ON，USE_MKL = ON，这是怎么回事？
'USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, '
USE_MKLDNN = ON，USE_MPI = OFF，USE_NCCL = ON，这是怎么回事？
'USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, '
USE_NNPACK = ON，USE_OPENMP = ON，USE_ROCM = OFF，使用：
'USE_ROCM_KERNEL_ASSERT=OFF, \n',
'USE_ROCM_KERNEL_ASSERT=OFF，\n'，
'Python': '3.10.14 (main, May 6 2024, 19:42:50) [GCC 11.2.0]',
'Python'：'3.10.14（main，May 6 2024，19：42：50）[GCC 11.2.0'，
'TorchVision': '0.18.0',
'TorchVision'：'0.18.0'，
'numpy_random_seed': 2147483648,
'numpy_random_seed'：2147483648，
'opencompass': '0.2.5+d656e81',
'opencompass'：'0.2.5+ d656e81'，
'sys.platform': 'linux'}
"sys. platform"："linux"}

Reproduces the problem - code/configuration sample

Attempt to modify the "run(self)" function in tools/case_analyzer.py

    def run(self):
        filename = get_infer_output_path(
            self.model_cfg, self.dataset_cfg,
            osp.join(self.work_dir, 'predictions')
        )
        root, ext = osp.splitext(filename)
        partial_filename = root + '_0' + ext
        if not osp.exists(osp.realpath(filename)) and not osp.exists(
                osp.realpath(partial_filename)):
            print(f'{filename} not found')
            return
        dataset = build_dataset_from_cfg(self.dataset_cfg)

        print("==========       build-dataset       ========")
        print(dataset)

        if 'dataset_postprocessor' in self.eval_cfg:
            def postprocess(sample):
                s = sample[self.ds_column] 
                proc = TEXT_POSTPROCESSORS.get(
                    self.eval_cfg['dataset_postprocessor']['type']
                )
                sample[self.ds_column] = proc(s)
                print("==========   sample ds_column    ========")
                print(sample)
                print("self.ds_column")
                return sample
            dataset = dataset.map(postprocess)
            print("==========   postprocess dataset    ========")
            print(dataset)

        if osp.exists(osp.realpath(filename)):
            preds = mmengine.load(filename)
        else:
            filename = partial_filename
            preds, offset = {}, 0
            i = 1
            while osp.exists(osp.realpath(filename)):
                _preds = mmengine.load(filename)  
                filename = root + f'_{i}' + ext
                i += 1
                for _o in range(len(_preds)):
                    preds[str(offset)] = _preds[str(_o)]
                    offset += 1
        pred_strs = [preds[str(i)]['prediction'] for i in range(len(preds))]
        if 'in-context examples' in preds['0']:
            print("___________   pred_strs   ↓ ________")
            print(pred_strs)
            print("___________   pred_strs   ↑ ________")
        if 'pred_postprocessor' in self.eval_cfg:
            print("===========    qwq     ==============")
            proc = TEXT_POSTPROCESSORS.get(
                self.eval_cfg['pred_postprocessor']['type']
            )
            pred_strs = [proc(s) for s in pred_strs]
#        if self.ds_split:
#            references = dataset[self.ds_split][self.ds_column]
#        else:
#            references = dataset[self.ds_column]

#        if len(pred_strs) != len(references):
#            print('length mismatch')
#            return
        allcase, badcase = [], []
        print("===========    gen origin_prompt    ===============")
        if 'in-context examples' in preds['0']:
            for i, (pred_str) in enumerate(zip(tqdm(pred_strs))):
                #ref_str = str(reference)
                ref_str = str(i)
#                print("===========    gen origin_prompt    ===============")
#                print("i:        ",i,"ref_str:     ",ref_str)
#                print("")    
                try:
                    pred_prompt = preds[str(i)]['label: ' + str(pred_str)]['testing input']
                    pred_PPL = preds[str(i)]['label: ' + str(pred_str)]['PPL']
                    print("===========    gen origin_prompt    ===============")
                    print(pred_prompt)
                    print("PPL:                   ",pred_PPL)
                    print("")

                    ref_prompt = preds[str(i)]['label: ' + ref_str]['testing input']
                    ref_PPL = preds[str(i)]['label: ' + ref_str]['PPL']
                except KeyError:
                    continue
                item = {
                    'prediction_prompt': pred_prompt,
                    'prediction': pred_str,
                    'prediction_PPL': pred_PPL,
                    'reference_prompt': ref_prompt,
                    'reference': ref_str,
                    'reference_PPL': ref_PPL
                }
                if pred_str != ref_str:
                    badcase.append(item)
                    allcase.append(item)
                else:
                    allcase.append(item)
        else:
            print("===========    gen pred_str    ===============")
            pred_str = zip(tqdm(pred_strs))
#            print(pred_str)
#            print("")
#            print(reference)
#            print("")
            for i, (pred_str) in enumerate(zip(tqdm(pred_strs))):
                #ref_str = str(reference)
                origin_prompt = preds[str(i)]['origin_prompt']
#                print("===========    gen origin_prompt    ===============")
#                print(origin_prompt)
#                print("")
                
                item = {
                    'origin_prompt': origin_prompt,
                    'prediction': pred_str,
                    #'reference': ref_str
                }
                badcase.append(item)
                allcase.append(item)
        out_path = get_infer_output_path(
            self.cfg['model'], self.cfg['dataset'],
            osp.join(self.work_dir, 'case_analysis/bad')
        )
        mkdir_or_exist(osp.split(out_path)[0])
        with open(out_path, 'w', encoding='utf-8') as f:
            json.dump(badcase, f, indent=4, ensure_ascii=False)

#        out_path = get_infer_output_path(
#            self.cfg['model'], self.cfg['dataset'],
#            osp.join(self.work_dir, 'case_analysis/all')
#        )
#        mkdir_or_exist(osp.split(out_path)[0])

        with open(out_path, 'w', encoding='utf-8') as f:
            json.dump(allcase, f, indent=4, ensure_ascii=False)

Reproduces the problem - command or script

python tools/case_analyzer.py configs/eval_demo.py

Reproduces the problem - error message

/home/qwq/anaconda3/envs/opencompass/lib/python3.10/site-packages/datasets/load.py:1486: FutureWarning: The repository for winograd_wsc contains custom code which must be executed to correctly load the dataset. You can inspect the repository content at https://hf.co/datasets/winograd_wsc
You can avoid this message in future by passing the argument trust_remote_code=True.
Passing trust_remote_code=True will be mandatory to load this dataset from the next major release of datasets.
warnings.warn(

Other information

无法读取ppl类数据集内容，但gen类正常